Welcome!
Thank you for viewing an example report created for the Effective Statistician conference.
Clients come to Exploristics with highly bespoke issues that require innovative and novel clinical trial designs to achieve their goals. Our Statistical Consultants need to produce bespoke data visualisations and reports. The Virtual Population Visualization application was designed to provide a flexible approach for creating data visualisations whilst streamlining workflows and producing consistent & professional plots.
Example Project
This is an example of a Vaccine trial, across several countries, low-incidence rate - so need to recruit large number of patients across multiple countries.
Using KerusCloud we simulated Virtual Populations with recruitment sites in 4 different countries, baseline patient data and repeated measure & time to event variables of disease occurrence.
Patient Recruitment
In this particular clinical trial example patient recruitment is a a key factor to the success or failure. With a rare disease, the vaccine will need to be administered to a large number of people and potentially on a longer duration.
Optimising the clinical trial design whilst considering patient recruitment, duration in addition to statistical properties increases the likelihood of success, reduces the cost, and gets a vaccine to patients that urgently need it sooner - saving lives!
Recruitment Sites
In this example we have 4 recruitment sites in 4 different countries in Africa. Maps are always a good introductory visualisation as they set the geographical context of a project and provide an indication of the scope of a project.
Code
# Load the library.
library(leaflet)
# Make a map
leaflet() %>%
addTiles() %>% # Add default OpenStreetMap map tiles
addMarkers(lng = c(9.0820, 30.55, 7.54, 17.607), lat = c(8.6753, -22.93, 5.554, -8.08), popup = c("Site A", "Site B", "Site C", "Site D"))Recruitment Profile
A waffle plot can tell us how many patients we’re recruiting from individual sites which is useful for logistical planning.
Including time into our recruitment profile allows us to know how long a clinical trial will take and allowing planning of interim analyses or alternative study designs such as group sequential.
Timeline
A visualisation such as a ganntt chart can provide a excellent summary of the project. It provides a context of the expected time of the trial and allows identification of any operational issues that may arise.
Simulating Data
KerusCloud generates simulated data based up to 17 difference distribution types and allows distribution parameters to be defined in strafed approach, allowing highly precise representations of data.